Xuzeng He

Latest effort in assessing the security of the code generated by large
language models

<figure>
<img src="https://cdn-images-1.medium.com/max/894/0*FN7wsPz9-E8BDA0Q" />
<figcaption>Source: Generated using DeviantArt’s DreamUp</figcaption>
</figure>

### Author

· Xuzeng He (**ORCID:**
[0009--0005--7317--7426](https://orcid.org/0009-0005-7317-7426))

### Introduction

With the surge of Large Language Models (LLMs) nowadays, there is a
rising trend among developers to use Large Language Models to assist
their daily code writing. Famous products include GitHub Copilot or
simply ChatGPT. However, just like those codes written by human
developers, code generated by LLMs can sometimes have a lot of security
risks as well. For example, there is a popular one called
"**Out-of-bounds Write**", which may allow attackers to write malicious
information into the computer memory for some potential criminal
activities.

In this post, we introduce some of the latest efforts in assessing the
security of the code generated by Large Language Models so that one can
have some basic ideas about how unreliable such models can be in general
cases or when given some specific tasks.

### Systematic experiment through CWEs

[Common Weakness
Enumeration](https://cwe.mitre.org/about/new_to_cwe.html) (CWE) is a
system operated by
[MITRE](https://en.wikipedia.org/wiki/Mitre_Corporation) that
categorises security vulnerabilities, including over 400 categories
regarding software security risks. For example, the security risk
"Out-of-bounds Write" mentioned earlier is assigned to "CWE-787". By
using CWE as the metric for evaluating the security risk of the
generated code, one can easily analyse a group of security
vulnerabilities in a systematic way.

<figure>
<img
src="https://cdn-images-1.medium.com/max/1024/0*7u5ddbb3_WlpVbfb" />
<figcaption>2023 CWE Top 25 Most Dangerous Software Weakness Link: <a
href="https://cwe.mitre.org/top25/archive/2023/2023_top25_list.html">https://cwe.mitre.org/top25/archive/2023/2023_top25_list.html</a></figcaption>
</figure>

In a recent research, to gain insight into the question "Are GitHub
Copilot's suggestions commonly insecure", [GitHub
Copilot's](https://github.com/features/copilot?ef_id=_k_CjwKCAjwxLKxBhA7EiwAXO0R0Gs52HhylRUPoX6UrxZB9nTpN40tyKrMPLApdsZfTdj3C3pkoFi69RoCMfoQAvD_BwE_k_&OCID=AIDcmm7mj0sudh_SEM__k_CjwKCAjwxLKxBhA7EiwAXO0R0Gs52HhylRUPoX6UrxZB9nTpN40tyKrMPLApdsZfTdj3C3pkoFi69RoCMfoQAvD_BwE_k_&gad_source=1&gclid=CjwKCAjwxLKxBhA7EiwAXO0R0Gs52HhylRUPoX6UrxZB9nTpN40tyKrMPLApdsZfTdj3C3pkoFi69RoCMfoQAvD_BwE)
generated code was evaluated regarding a subset of MITRE's CWEs. To be
more specific, Copilot's behaviour is studied along three dimensions,
including diversity of weakness, diversity of prompt and diversity of
domain. **GitHub Copilot** is a code completion tool developed by
[GitHub](https://en.wikipedia.org/wiki/GitHub) and
[OpenAI](https://en.wikipedia.org/wiki/OpenAI) that assists users of
[Visual Studio Code](https://en.wikipedia.org/wiki/Visual_Studio_Code),
[Visual Studio](https://en.wikipedia.org/wiki/Visual_Studio),
[Neovim](https://en.wikipedia.org/wiki/Neovim), and
[JetBrains](https://en.wikipedia.org/wiki/JetBrains) [integrated
development
environments](https://en.wikipedia.org/wiki/Integrated_development_environment)
(IDEs) by autocompleting code.

For the **diversity of weakness**, three different scenarios for each
applicable "top 25" CWEs were constructed as small, incomplete programs
waiting to be filled out by Copilot. As code evaluation,
[CodeQL](https://codeql.github.com/) software scanning was used along
with manual inspection to assess whether the suggestions returned were
vulnerable to that CWE.

Next, for the **diversity of prompt**, the effect that different prompts
(aka the input we give to the model) have on how likely Copilot is to
return suggestions that are vulnerable to a selected CWE (e.g. SQL
injection) are assessed.

Eventually, for the **diversity of domain**, the security of the code
generated by Copilot is also assessed by focusing on the Copilot's
behaviour when it is tasked with a new domain added to CWEs, such as
code written in another programming language.

In this case, Copilot's response to these scenarios is mixed from a
security standpoint, where around 40% of the generation results are
found to be vulnerable.

### User Studies

Apart from general systematic testing, user studies were also actively
conducted to assess the security of the generated code. Recently, a
security-driven user study to assess code written by student programmers
when assisted by LLMs was conducted.

Students for this study were recruited through social media and split
randomly into 'control' (no LLM access) and 'assisted' (with LLM access)
groups. They were prompted to complete a shopping list program
implementation in C. C was chosen here since a majority of security
risks are memory-based issues in low level languages such as C/C++).

An LLM called code-cushman-001 was used in this study to help the
students in the *assisted group* due to its fast speed to operate and
its similar response time to GitHub Copilot. After generating the
function, compilation would be checked and another completion would be
requested if the result fails. To analyse the results, [standard
statistical hypothesis
tests](https://www.simplilearn.com/tutorials/statistics-tutorial/hypothesis-testing-in-statistics)
were used to provide evidence-based conclusions. Hypothesis Testing is a
type of statistical analysis in which assumptions about a population
parameter are put to the test. It is used to estimate the relationship
between 2 statistical variables.

By examining the completed code for functionality and security, using
manual and automated methods regarding a list of CWEs, the study found
that the security impacts are minimal in this setting of using LLMs for
code generation assistance. Existing findings on the productivity
benefits of AI assistance were confirmed. They also found that the
AI-assisted group produced security-critical bugs at a rate no greater
than 10% higher than the control group (non-assisted). When
investigating the origin of bugs within the assisted users, 63% of the
bugs originated in code written by humans and 36% of the bugs were
present in the taken suggestions.

### Conclusion

The use of Large Language Models has witnessed significant advancement
in multiple directions, especially when they are used in assisting code
implementation. However, code generated by Large Language Models can
have some potential security risks. In this post, we discussed the
latest effort in assessing the security of the code generated by large
language models, ranging from systematic testing to user studies.
Through continuous investigation and refinement, we believe that the use
of Large Language Models can open up exciting opportunities for us in
code generation.

### References

- Pearce, H.; Ahmad, B.; Tan, B.; Dolan-Gavitt, B.; and Karri, R., 2021.
  Asleep at the keyboard? assessing the security of github copilot's
  code contributions. <https://arxiv.org/abs/2108.09293>
- Sandoval, G.; Pearce, H.; Nys, T.; Karri, R.; Garg, S.; and
  Dolan-Gavitt, B., 2022. Lost at c: A user study on the security
  implications of large language model code assistants.
  <https://arxiv.org/abs/2208.09727>

![](https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f305f9f01ce9){width="1"
height="1"}


Latest effort in assessing the security of the code generated by large language models Author   · Xuzeng He (
<strong>
 ORCID:
</strong>
0009–0005–7317–7426) Introduction   With the surge of Large Language Models (LLMs) nowadays, there is a rising trend among developers to use Large Language Models to assist their daily code writing. Famous products include GitHub Copilot or simply ChatGPT.

Messaggi di Rogue Scholar

Large Language Models for Code Writing: Security Assessment